This dissertation is about the institutional choices governments make to manage their petroleum wealth. It is about the determinants of these choices, but more importantly, their consequences for effective governance and how they explain variations in political outcomes in oil-producing countries. I begin by describing several different institutional pathways -- involving national oil companies (NOCs) and their varying characteristics -- that governments can take in extracting petroleum and regulating its production. My goal, then, is to show how these seemingly technical institutional choices can have profound impacts on governance, ranging from effects on state revenue collection to incentives for corruption to ultimately the survival of the regime itself.
To this aim, I collected original longitudinal data on the formation of NOCs in 62 countries since 1900; data from U.S. Department of Justice transcripts on the prosecution of corrupt practices in the energy sectors of 80 countries in the 2006-12 period; and existing cross-national data on government revenue capture from the sale of oil and natural resources. I analyze the determinants of NOC formation in the first empirical chapter, where I use Bayesian analysis informed with interview-based data from oil consultants to test and confirm leading theories of state revenue-maximization as the primary determinant of expropriation. In the following chapter, I analyze the process of extortion in the oil sector where I show cross-nationally how NOC institutional design influences bribery to high-level government officials. In the penultimate chapter, I expand on the governance consequences of NOCs by showing that nationalization ultimately increases state resource revenues, creating pathways for regime stability and duration. In the last chapter, I discuss the theoretical implications of my argument and findings.
I make two broad claims in this dissertation. First, while there is much agreement that oil is not always and everywhere a curse for political development, there is little consensus about the specific conditions or institutions that do and do not matter. I help turn the ``institutions matter'' phrase from a vague stylized fact into a well-measured, clearly-specified phenomenon. Second, when it comes to the production of natural resources, classical economics theories would suggest that state intervention will lead to inefficiency, lower outputs and therefore lower revenues. In contrast, I argue and show evidence that some forms of state intervention -- that is, certain types of NOCs -- actually increase both production levels and revenues when compared to periods of no state intervention. Taken together, my dissertation applies novel ways to measure and test theories about oil's conditional effects on politics that are widely circulated but often assumed rather than tested.